Predicting Early Death Among Elderly Dialysis Patients: Development and Validation of a Risk Score to Assist Shared Decision Making for Dialysis Initiation

PTH may predict early mortality in incident patients on hemodialysis: results from a large cohort

PURPOSE

Parathyroid hormone (PTH) is merit as a risk factor for mortality in patients with chronic kidney disease in prevalent hemodialysis patients in a U shape. Most studies, however, do not focus on incident patients and those who died within the first 90 days of therapy. We evaluated PTH as a risk factor for mortality in a large cohort population in Brazil.

METHODS

This is an observational cohort study that included 4317 adult patients who initiated hemodialysis between July 1st, 2012 and June 30, 2017. The main outcome was all-cause mortality. Fine-gray sub-distribution hazard models were used to evaluate survival in the presence of a competing event (kidney transplant).

RESULTS

Median PTH levels of 252 (118, 479) pg/mL. There were 331 deaths during the first 90 days of therapy (6.7%), 430 in a 1-year follow-up (10.7%) and 1282 (32%) during the 5-year study period. Deaths according to PTH < 150, 150-600 and > 600 pg/mL corresponded to 38.1%, 33.0% and 28.5%, respectively (p < 0.001). In an adjusted model, patients who started dialysis with PTH < 150 pg/mL had a higher mortality risk within the first 90 days, but not in 1 year and 5 years after starting dialysis. Analyses in a subset of patients with a repeated PTH in 1 year (N = 1954) showed that although persistent PTH low levels (< 150 pg/mL) at 1 year were significantly associated with all-cause mortality, this result was not sustained after multiple adjustments.

CONCLUSION

PTH < 150 pg/mL confers a high mortality risk in the first 90 days of dialysis. If this result reflects poor nutritional conditions, it deserves further investigations.

External validation of the 2-year mortality prediction tool in hemodialysis patients developed using a Bayesian network

Background

In recent years, a number of predictive models have appeared to predict the risk of medium-term mortality in hemodialysis patients, but only one, limited to patients aged over 70 years, has undergone sufficiently powerful external validation. Recently, using a national learning database and an innovative approach based on Bayesian networks and 14 carefully selected predictors, we have developed a clinical prediction tool to predict all-cause mortality at 2 years in all incident hemodialysis patients. In order to generalize the results of this tool and propose its use in routine clinical practice, we carried out an external validation using an independent external validation database.

Methods

A regional, multicenter, observational, retrospective cohort study was conducted to externally validate the tool for predicting 2-year all-cause mortality in incident and prevalent hemodialysis patients. This study recruited a total of 142 incident and 697 prevalent adult hemodialysis patients followed up in one of the eight Association pour l'Utilisation du Rein Artificiel dans la région Lyonnaise (AURAL) Alsace dialysis centers.

Results

In incident patients, the 2-year all-cause mortality prediction tool had an area under the receiver curve (AUC-ROC) of 0.73, an accuracy of 65%, a sensitivity of 71% and a specificity of 63%. In prevalent patients, the performance for the external validation were similar in terms of AUC-ROC, accuracy and specificity, but was lower in term of sensitivity.

Conclusion

The tool for predicting all-cause mortality at 2 years, developed using a Bayesian network and 14 routinely available explanatory variables, obtained satisfactory external validation in incident patients, but sensitivity was insufficient in prevalent patients.

Development of a prognostic risk score to predict early mortality in incident elderly Japanese hemodialysis patients

BACKGROUND

Information of short-term prognosis after hemodialysis (HD) introduction is important for elderly patients with chronic kidney disease (CKD) and their families choosing a modality of renal replacement therapy. Therefore, we developed a risk score to predict early mortality in incident elderly Japanese hemodialysis patients.

MATERIALS AND METHODS

We analyzed data of incident elderly HD patients from a nationwide cohort study of the Japanese Society for Dialysis Therapy Renal Data Registry (JRDR) to develop a prognostic risk score. Candidate risk factors for early death within 1 year was evaluated using multivariate logistic regression analysis. The risk score was developed by summing up points derived from parameter estimate values of independent risk factors. The association between risk score and early death was tested using Cox proportional hazards models. This risk score was validated twice by using an internal validation cohort derived from the JRDR and an external validation cohort collected for this study.

RESULTS

Using the development cohort (n = 2,000), nine risk factors were retained in the risk score: older age (>85), yes = 2, no = 0; sex, male = 2, female = 0; lower body mass index (<20), yes = 2, no = 0; cancer, yes = 1, no = 0; dementia, yes = 3, no = 0; lower creatinine (<6.5 mg/dL), yes = 1, no = 0; lower albumin (<3.0 g/dL), yes = 3, no = 0; normal or high calcium (≥8.5 mg/dL), yes = 1, no = 0; and higher C reactive protein (>2.0 mg/dL), yes = 2, no = 0. In the internal and external validation cohorts (n = 739, 140, respectively), the medium- and high-risk groups (total score, 6 to 10 and 11 or more, respectively) showed significantly higher risk of early death than the low-risk group (total score, 0 to 5) (p<0.001).

CONCLUSION

We developed a prognostic risk score predicting early death within 1 year in incident elderly Japanese HD patients, which may help detect elderly patients with a high-risk of early death after HD introduction.

The estimated mediating roles of anemia-related variables in the association between kidney function and mortality: a National Health and Nutrition Examination Survey (NHANES) study

Anemia is a common complication of chronic kidney disease (CKD), impacting long-term outcomes such as mortality and morbidity. Analyzing NHANES data from 1999 through 2016 for adults aged ≥ 20 years, we assessed the mediating effects of anemia biomarkers (hemoglobin, hematocrit, red cell distribution width [RDW], and mean corpuscular hemoglobin concentration [MCHC]) on CKD-related outcomes by using hazard ratios from a biomarker-adjusted model. Of 44,099 participants, 7463 experienced all-cause death. Cox proportional hazard models revealed a higher all-cause mortality risk in the > 45 years and CKD groups than in the early CKD group. Hemoglobin, hematocrit and MCHC were inversely related to all-cause mortality; RDW was related to mortality. Single mediation analysis showed greater mediating effects of anemia indicators on CKD and mortality in the elderly (> 65 years) population than those in the general population. In the multimediation analysis, the combined mediating effect of anemia was higher in the CKD population than in the general population. This study showed a proportional increase in the mediating effect of anemia with CKD stage, suggesting potential therapeutic avenues. However, further exploration of other mediating factors on kidney outcomes is necessary.

Importance of dialysis specialists in early mortality in elderly hemodialysis patients: a multicenter retrospective cohort study

The early mortality rate in elderly patients undergoing hemodialysis is more than twice that in young patients, requiring more specialized healthcare. We investigated whether the number of professional dialysis specialists affected early mortality in elderly patients undergoing hemodialysis. This multicenter retrospective cohort study analyzed data from 1860 patients aged ≥ 70 years who started hemodialysis between January 2010 and December 2017. Study regions included Seoul, Gyeonggi-do, Gangwon-do, Daejeon/Chungcheong-do, Daegu/Gyeongsangbuk-do, and Busan/Ulsan/Gyeongsangnam-do. The number of patients undergoing hemodialysis per dialysis specialist was calculated using registered data from each hemodialysis center. Early mortality was defined as death within 6 months of hemodialysis initiation. Gangwon-do (28.3%) and Seoul (14.5%) showed the highest and lowest early mortality rate, respectively. Similarly, Gangwon-do (64.6) and Seoul (43.9) had the highest and lowest number of patients per dialysis specialist, respectively. Relatively consistent results were observed for the regional rankings of early mortality rate and number of patients per dialysis specialist. Multivariate Cox regression analysis-adjusted for previously known significant risk factors-revealed that the number of patients per dialysis specialist was an independent risk factor for early mortality (hazard ratio: 1.031, p < 0.001). This study underscores the growing need for dialysis specialists for elderly hemodialysis patients in Korea.

Discrepancy in responses to the surprise question between hemodialysis nurses and physicians, with focus on patient clinical characteristics: A comparative study

INTRODUCTION

The surprise question (SQ) "Would I be surprised if this patient died within the next xx months" can be used by different professions to foresee the need of serious illness conversations in patients approaching end of life. However, little is known about the different perspectives of nurses and physicians in responses to the SQ and factors influencing their appraisals. The aim was to explore nurses' and physicians' responses to the SQ regarding patients on hemodialysis, and to investigate how these answers were associated with patient clinical characteristics.

METHODS

This comparative cross-sectional study included 361 patients for whom 112 nurses and 15 physicians responded to the SQ regarding 6 and 12 months. Patient characteristics, performance status, and comorbidities were obtained. Cohen's kappa was used to analyze the interrater agreement between nurses and physicians in their responses to the SQ and multivariable logistic regression was applied to reveal the independent association to patient clinical characteristics.

FINDINGS

Proportions of nurses and physicians responding to the SQ with "no, not surprised" was similar regarding 6 and 12 months. However, there was a substantial difference concerning which specific patient the nurses and physicians responded "no, not surprised", within 6 (κ = 0.366, p < 0.001, 95% CI = 0.288-0.474) and 12 months (κ = 0.379, p < 0.001, 95% CI = 0.281-0.477). There were also differences in the patient clinical characteristics associated with nurses' and physicians' responses to the SQ.

DISCUSSION

Nurses and physicians have different perspectives in their appraisal when responding to the SQ for patients on hemodialysis. This may reinforce the need for communication and discussion between nurses and physicians to identify the need of serious illness conversations in patients approaching the end of life, in order to adapt hemodialysis care to patient preferences and needs.

Predicting outcomes in chronic kidney disease: needs and preferences of patients and nephrologists

INTRODUCTION

Guidelines on chronic kidney disease (CKD) recommend that nephrologists use clinical prediction models (CPMs). However, the actual use of CPMs seems limited in clinical practice. We conducted a national survey study to evaluate: 1) to what extent CPMs are used in Dutch CKD practice, 2) patients' and nephrologists' needs and preferences regarding predictions in CKD, and 3) determinants that may affect the adoption of CPMs in clinical practice.

METHODS

We conducted semi-structured interviews with CKD patients to inform the development of two online surveys; one for CKD patients and one for nephrologists. Survey participants were recruited through the Dutch Kidney Patient Association and the Dutch Federation of Nephrology.

RESULTS

A total of 126 patients and 50 nephrologists responded to the surveys. Most patients (89%) reported they had discussed predictions with their nephrologists. They most frequently discussed predictions regarded CKD progression: when they were expected to need kidney replacement therapy (KRT) (n = 81), and how rapidly their kidney function was expected to decline (n = 68). Half of the nephrologists (52%) reported to use CPMs in clinical practice, in particular CPMs predicting the risk of cardiovascular disease. Almost all nephrologists (98%) reported discussing expected CKD trajectories with their patients; even those that did not use CPMs (42%). The majority of patients (61%) and nephrologists (84%) chose a CPM predicting when patients would need KRT in the future as the most important prediction. However, a small portion of patients indicated they did not want to be informed on predictions regarding CKD progression at all (10-15%). Nephrologists not using CPMs (42%) reported they did not know CPMs they could use or felt that they had insufficient knowledge regarding CPMs. According to the nephrologists, the most important determinants for the adoption of CPMs in clinical practice were: 1) understandability for patients, 2) integration as standard of care, 3) the clinical relevance.

CONCLUSION

Even though the majority of patients in Dutch CKD practice reported discussing predictions with their nephrologists, CPMs are infrequently used for this purpose. Both patients and nephrologists considered a CPM predicting CKD progression most important to discuss. Increasing awareness about existing CPMs that predict CKD progression may result in increased adoption in clinical practice. When using CPMs regarding CKD progression, nephrologists should ask whether patients want to hear predictions beforehand, since individual patients' preferences vary.

Survival Analysis and Its Predictors Among Hemodialysis Patients at Saint Paul Hospital Millennium Medical College and Myungsung Christian Medical Center in Addis Ababa, Ethiopia, 2021

Background

Diabetes mellitus and hypertension are the most prominent conditions causing chronic kidney disease and eventually end-stage renal disease. Renal replacement therapy, particularly hemodialysis (HD), is the mainstay of treatment. The aim of this study is to assess the overall survival status of HD patients and potential survival predictors at Saint Paul hospital millennium medical college (SPHMMC) and Myungsung Christian Medical Center (MCM) in Addis Ababa, Ethiopia.

Methods

A retrospective cohort study was conducted on HD patients at SPHMMC and MCM general hospital from January 1, 2013 to December 30, 2020. Kaplan Meier, Log-rank, and Cox proportional regression models were used for the analysis. Estimated risks were reported as hazard ratios with 95% confidence intervals and P<0.05 was considered as having a significant association.

Results

A total of 128 patients were included in the study. Median survival time was 65 months. The predominant co-morbid condition was found to be diabetes mellitus with hypertension (42%). The total risk time for these patients was 143,617 person years. The overall incidence rate of death was 2.9 per 10,000 person years (95% CI=2.2-4). Patients who developed blood stream infection were 2.98-times more likely to die than those without infection. Those using an arteriovenous fistula were 66% less likely to die than those using a central venous catheter. Additionally, patients treated in a government-owned facility were 79% less likely to die.

Conclusion

The study identified that the median survival time of 65 months was comparable with developed nations. Significant predictors of death were found to be blood stream infection and type of vascular access. Government-owned treatment facilities showed better patient survival.

Treatment decision-making and care among older adults with kidney failure: protocol for a multicentre, prospective observational cohort study with nested substudies and linked qualitative research (the Elderly Advanced CKD Programme)

INTRODUCTION

Shared treatment decision-making and planning of care are fundamental in advanced chronic kidney disease (CKD) management. There are limited data on several key outcomes for the elderly population including survival, quality of life, symptom burden, changes in physical functioning and experienced burden of healthcare. Patients, caregivers and clinicians consequently face significant uncertainty when making life-impacting treatment decisions. The Elderly Advanced CKD Programme includes quantitative and qualitative studies to better address challenges in treatment decision-making and planning of care among this increasingly prevalent elderly cohort.

METHODS AND ANALYSIS

The primary component is OUTcomes of Older patients with Kidney failure (OUTLOOK), a multicentre prospective observational cohort study that will enrol 800 patients ≥75 years with kidney failure (estimated glomerular filtration rate ≤15 mL/min/1.73 m2) across a minimum of six sites in Australia. Patients entered are in the decision-making phase or have recently made a decision on preferred treatment (dialysis, conservative kidney management or undecided). Patients will be prospectively followed until death or a maximum of 4 years, with the primary outcome being survival. Secondary outcomes are receipt of short-term acute dialysis, receipt of long-term maintenance dialysis, changes in biochemistry and end-of-life care characteristics. Data will be used to formulate a risk prediction tool applicable for use in the decision-making phase. The nested substudies Treatment modalities for the InfirM ElderLY with end stage kidney disease (TIMELY) and Caregivers of The InfirM ElderLY with end stage kidney disease (Co-TIMELY) will longitudinally assess quality of life, symptom burden and caregiver burden among 150 patients and 100 caregivers, respectively. CONsumer views of Treatment options for Elderly patieNts with kiDney failure (CONTEND) is an additional qualitative study that will enrol a minimum of 20 patients and 20 caregivers to explore experiences of treatment decision-making and care.

ETHICS AND DISSEMINATION

Ethics approval was obtained through Sydney Local Health District Human Research Ethics Committee (2019/ETH07718, 2020/ETH02226, 2021/ETH01020, 2019/ETH07783). OUTLOOK is approved to have waiver of individual patient consent. TIMELY, Co-TIMELY and CONTEND participants will provide written informed consent. Final results will be disseminated through peer-reviewed journals and presented at scientific meetings.

Development of a prediction model to estimate the 5-year risk of cardiovascular events and all-cause mortality in haemodialysis patients: a retrospective study

Background

Cardiovascular disease (CVD) is a major cause of mortality in patients on haemodialysis. The development of a prediction model for CVD risk is necessary to help make clinical decisions for haemodialysis patients. This retrospective study aimed to develop a prediction model for the 5-year risk of CV events and all-cause mortality in haemodialysis patients in China.

Methods

We retrospectively enrolled 398 haemodialysis patients who underwent dialysis at the dialysis facility of the General Hospital of Northern Theater Command in June 2016 and were followed up for 5 years. The composite outcome was defined as CV events and/or all-cause death. Multivariable logistic regression with backwards stepwise selection was used to develop our new prediction model.

Results

Seven predictors were included in the final model: age, male sex, diabetes, history of CV events, no arteriovenous fistula at dialysis initiation, a monocyte/lymphocyte ratio greater than 0.43 and a serum uric acid level less than 436 mmol/L. Discrimination and calibration were satisfactory, with a C-statistic above 0.80. The predictors lay nearly on the 45-degree line for agreement with the outcome in the calibration plot. A simple clinical score was constructed to provide the probability of 5-year CV events or all-cause mortality. Bootstrapping validation showed that the new model also has similar discrimination and calibration. Compared with the Framingham risk score (FRS) and a similar model, our model showed better performance.

Conclusion

This prognostic model can be used to predict the long-term risk of CV events and all-cause mortality in haemodialysis patients. An MLR greater than 0.43 is an important prognostic factor.

A Machine Learning Model for Predicting Mortality within 90 Days of Dialysis Initiation

Background

The first 90 days after dialysis initiation are associated with high morbidity and mortality in end-stage kidney disease (ESKD) patients. A machine learning-based tool for predicting mortality could inform patient-clinician shared decision making on whether to initiate dialysis or pursue medical management. We used the eXtreme Gradient Boosting (XGBoost) algorithm to predict mortality in the first 90 days after dialysis initiation in a nationally representative population from the United States Renal Data System.

Methods

A cohort of adults initiating dialysis between 2008-2017 were studied for outcome of death within 90 days of dialysis initiation. The study dataset included 188 candidate predictors prognostic of early mortality that were known on or before the first day of dialysis and was partitioned into training (70%) and testing (30%) subsets. XGBoost modeling used a complete-case set and a dataset obtained from multiple imputation. Model performance was evaluated by c-statistics overall and stratified by subgroups of age, sex, race, and dialysis modality.

Results

The analysis included 1,150,195 patients with ESKD, of whom 86,083 (8%) died in the first 90 days after dialysis initiation. The XGBoost models discriminated mortality risk in the nonimputed (c=0.826, 95% CI, 0.823 to 0.828) and imputed (c=0.827, 95% CI, 0.823 to 0.827) models and performed well across nearly every subgroup (race, age, sex, and dialysis modality) evaluated (c>0.75). Across predicted risk thresholds of 10%-50%, higher risk thresholds showed declining sensitivity (0.69-0.04) with improving specificity (0.79-0.99); similarly, positive likelihood ratio was highest at the 40% threshold, whereas the negative likelihood ratio was lowest at the 10% threshold. After calibration using isotonic regression, the model accurately estimated the probability of mortality across all ranges of predicted risk.

Conclusions

The XGBoost-based model developed in this study discriminated risk of early mortality after dialysis initiation with excellent calibration and performed well across key subgroups.

Shared decision-making in advanced kidney disease: a scoping review

OBJECTIVES

To provide a comprehensive overview of interventions that support shared decision-making (SDM) for treatment modality decisions in advanced kidney disease (AKD). To provide summarised information on their content, use and reported results. To provide an overview of interventions currently under development or investigation.

DESIGN

The JBI methodology for scoping reviews was followed. This review conforms to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) checklist.

DATA SOURCES

MEDLINE, Embase, Web of Science, Cochrane Library, Emcare, PsycINFO, PROSPERO and Academic Search Premier for peer-reviewed literature. Other online databases (eg, clinicaltrials.gov, OpenGrey) for grey literature.

ELIGIBILITY FOR INCLUSION

Records in English with a study population of patients >18 years of age with an estimated glomerular filtration rate <30 mL/min/1.73 m2. Records had to be on the subject of SDM, or explicitly mention that the intervention reported on could be used to support SDM for treatment modality decisions in AKD.

DATA EXTRACTION AND SYNTHESIS

Two reviewers independently screened and selected records for data extraction. Interventions were categorised as prognostic tools (PTs), educational programmes (EPs), patient decision aids (PtDAs) or multicomponent initiatives (MIs). Interventions were subsequently categorised based on the decisions they were developed to support.

RESULTS

One hundred forty-five interventions were identified in a total of 158 included records: 52 PTs, 51 EPs, 29 PtDAs and 13 MIs. Sixteen (n=16, 11%) were novel interventions currently under investigation. Forty-six (n=46, 35.7%) were reported to have been implemented in clinical practice. Sixty-seven (n=67, 51.9%) were evaluated for their effects on outcomes in the intended users.

CONCLUSION

There is no conclusive evidence on which intervention is the most efficacious in supporting SDM for treatment modality decisions in AKD. There is a lot of variation in selected outcomes, and the body of evidence is largely based on observational research. In addition, the effects of these interventions on SDM are under-reported.

Supporting Shared Decision-Making and Home Dialysis in End-Stage Kidney Disease

It has been widely demonstrated that patient education and empowerment, especially involving shared treatment decisions, improve patient outcomes in chronic medical conditions, including chronic kidney disease requiring kidney replacement therapies. Accordingly, regulatory agencies in the US and worldwide recommend shared decision-making for finalizing one's choice of kidney replacement therapy. It is also recognized that the US needs to substantially increase home dialysis utilization to leverage its positive impacts on patient and healthcare cost-related outcomes. This perspective highlights how the routine clinical use of the recommended practice of shared decision-making can exist in synergy with the system's goal for increased home dialysis use. It introduces a pragmatic provider checklist, The Nephrologist's Shared Decision-Making Checklist, grounded in the relevant theories of shared decision-making, and, unlike some research assessments and extant tools, is easy to understand and implement in clinical practice. This qualitative Checklist can help providers ensure that they have co-constructed an SDM experience with the patient and involved caretakers, helping them benefit from the improved outcomes associated with SDM.

Risk of bias of prognostic models developed using machine learning: a systematic review in oncology

BACKGROUND

Prognostic models are used widely in the oncology domain to guide medical decision-making. Little is known about the risk of bias of prognostic models developed using machine learning and the barriers to their clinical uptake in the oncology domain.

METHODS

We conducted a systematic review and searched MEDLINE and EMBASE databases for oncology-related studies developing a prognostic model using machine learning methods published between 01/01/2019 and 05/09/2019. The primary outcome was risk of bias, judged using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). We described risk of bias overall and for each domain, by development and validation analyses separately.

RESULTS

We included 62 publications (48 development-only; 14 development with validation). 152 models were developed across all publications and 37 models were validated. 84% (95% CI: 77 to 89) of developed models and 51% (95% CI: 35 to 67) of validated models were at overall high risk of bias. Bias introduced in the analysis was the largest contributor to the overall risk of bias judgement for model development and validation. 123 (81%, 95% CI: 73.8 to 86.4) developed models and 19 (51%, 95% CI: 35.1 to 67.3) validated models were at high risk of bias due to their analysis, mostly due to shortcomings in the analysis including insufficient sample size and split-sample internal validation.

CONCLUSIONS

The quality of machine learning based prognostic models in the oncology domain is poor and most models have a high risk of bias, contraindicating their use in clinical practice. Adherence to better standards is urgently needed, with a focus on sample size estimation and analysis methods, to improve the quality of these models.

Development and validation of a nomogram for predicting the 6-months survival rate of patients undergoing incident hemodialysis in China

Background

The all-cause mortality of patients undergoing hemodialysis (HD) is higher than in the general population. The first 6 months after dialysis are important for new patients. The aim of this study was to develop and validate a nomogram for predicting the 6-month survival rate of HD patients.

Methods

A prediction model was constructed using a training cohort of 679 HD patients. Multivariate Cox regression analyses were performed to identify predictive factors. The identified factors were used to establish a nomogram. The performance of the nomogram was assessed using the C-index and calibration plots. The nomogram was validated by performing discrimination and calibration tests on an additional cohort of 173 HD patients.

Results

During a follow-up period of six months, 47 and 16 deaths occurred in the training cohort and validation cohort, respectively, representing a mortality rate of 7.3% and 9.2%, respectively. The nomogram comprised five commonly available predictors: age, temporary dialysis catheter, intradialytic hypotension, use of ACEi or ARB, and use of loop diuretics. The nomogram showed good discrimination in the training cohort [C-index 0.775(0.693–0.857)] and validation cohort [C-index 0.758(0.677–0.836)], as well as good calibration, indicating that the performance of the nomogram was good. The total score point was then divided into two risk classifications: low risk (0–90 points) and high risk (≥ 91 points). Further analysis showed that all-cause mortality was significantly different between the high-risk group and the low-risk group.

Conclusions

The constructed nomogram accurately predicted the 6-month survival rate of HD patients, and thus it can be used in clinical decision-making.

Methodological conduct of prognostic prediction models developed using machine learning in oncology: a systematic review

BACKGROUND

Describe and evaluate the methodological conduct of prognostic prediction models developed using machine learning methods in oncology.

METHODS

We conducted a systematic review in MEDLINE and Embase between 01/01/2019 and 05/09/2019, for studies developing a prognostic prediction model using machine learning methods in oncology. We used the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement, Prediction model Risk Of Bias ASsessment Tool (PROBAST) and CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS) to assess the methodological conduct of included publications. Results were summarised by modelling type: regression-, non-regression-based and ensemble machine learning models.

RESULTS

Sixty-two publications met inclusion criteria developing 152 models across all publications. Forty-two models were regression-based, 71 were non-regression-based and 39 were ensemble models. A median of 647 individuals (IQR: 203 to 4059) and 195 events (IQR: 38 to 1269) were used for model development, and 553 individuals (IQR: 69 to 3069) and 50 events (IQR: 17.5 to 326.5) for model validation. A higher number of events per predictor was used for developing regression-based models (median: 8, IQR: 7.1 to 23.5), compared to alternative machine learning (median: 3.4, IQR: 1.1 to 19.1) and ensemble models (median: 1.7, IQR: 1.1 to 6). Sample size was rarely justified (n = 5/62; 8%). Some or all continuous predictors were categorised before modelling in 24 studies (39%). 46% (n = 24/62) of models reporting predictor selection before modelling used univariable analyses, and common method across all modelling types. Ten out of 24 models for time-to-event outcomes accounted for censoring (42%). A split sample approach was the most popular method for internal validation (n = 25/62, 40%). Calibration was reported in 11 studies. Less than half of models were reported or made available.

CONCLUSIONS

The methodological conduct of machine learning based clinical prediction models is poor. Guidance is urgently needed, with increased awareness and education of minimum prediction modelling standards. Particular focus is needed on sample size estimation, development and validation analysis methods, and ensuring the model is available for independent validation, to improve quality of machine learning based clinical prediction models.

Kidney Supportive Care in Peritoneal Dialysis: Developing a Person-Centered Kidney Disease Care Plan

Individuals receiving peritoneal dialysis (PD)—similar to those receiving hemodialysis —may experience high mortality coupled with a high symptom burden and reduced health-related quality of life. In this context, a discussion of the risks, benefits, and tradeoffs of PD and/or other kidney treatment modalities should be explored based on individual goals and preferences. Through these principles, kidney supportive care provides a person-centered approach to kidney disease care throughout the spectrum of kidney failure and earlier stages of chronic kidney disease. Kidney supportive care is offered in conjunction with life-prolonging therapies, including dialysis and kidney transplants, and is increasingly recognized as an integral part of advancing the care of PD patients. Using “My Kidney Care Roadmap” for shared decision making, kidney supportive care guides patients undergoing PD and their clinicians to (1) elicit patient goals, values, and priorities; (2) convey medical prognosis and suitable treatment options; and (3) ask “Which of these kidney treatment options will best help me achieve my goals and priorities?” to inform both current and future decisions, including choice of dialysis modalities, time-limited trials, and/or nondialysis management. Recognizing that patient priorities and choices may evolve, this framework ultimately allows patients to continually reassess their PD care to better achieve goal-directed dialysis.

Nierenersatzverfahren bei Hochbetagten

Hochbetagte haben an der Hämodialyse eine 1‑Jahres-Mortalität, die im Zusammenhang mit Komorbiditäten und einem Katheter als Dialysezugang 30 % übersteigt. Metaanalysen zeigen aber, dass frühzeitige Vorbereitung und individuelle Verfahrensauswahl die Morbidität und Mortalität auch im hohen Lebensalter entscheidend bessern. Mit zunehmendem Alter und Gebrechlichkeit verschieben sich dabei die Behandlungsziele weg von der Verlängerung der Lebensdauer auf die Verbesserung der Lebensqualität. Damit kann die Präferenz von Heimdialyseverfahren, auch als assistierte Peritonealdialyse, ebenso Bedeutung erringen wie die fachnephrologische Behandlung ohne Nierenersatzverfahren mit palliativem Therapieziel. Im höheren Lebensalter bestimmen zunehmend Komorbiditäten, kognitive Einschränkungen, Gebrechlichkeit und die Gesamtprognose das sinnvolle Vorgehen. Bereits bei der Anlage von Gefäßzugängen ergeben sich hinsichtlich Anastomosenort und Anlagezeitpunkt bei Hochbetagten andere Entscheidungskriterien. Empfehlungen zu Dialysedauer und -frequenz folgen der Lebensqualität mit inkrementellen und am Ende des Lebens auch dekrementellen Therapieregimen. Die demographische Entwicklung stellt die Nephrologie mit einer Zunahme älterer Patienten vor besondere Herausforderungen. Frühzeitige Aufklärung über alle Nierenersatzverfahren und die Festlegung individueller Therapieziele können bei sorgfältiger Auswahl von Dialysemodalität und -intensität auch bei Hochbetagten entscheidend zur Verbesserung der Prognose und insbesondere der Lebensqualität beitragen.

Communication principles and practices for making shared decisions about renal replacement therapy: a review of the literature

PURPOSE OF REVIEW

To provide an overview of the skill set required for communication and person-centered decision making for renal replacement therapy (RRT) choices, especially conservative kidney management (CKM).

RECENT FINDINGS

Research on communication and decision-making skills for shared RRT decision making is still in infancy. We adapt literature from other fields such as primary care and oncology for effective RRT decision making.

SUMMARY

We review seven key skills: (1) Announcing the need for decision making (2) Agenda Setting (3) Educating patients about RRT options (4) Discussing prognoses (5) Eliciting patient preferences (6) Responding to emotions and showing empathy, and (7) Investing in the end. We also provide example sentences to frame the conversations around RRT choices including CKM.

Predicting respiratory failure for COVID-19 patients in Japan: a simple clinical score for evaluating the need for hospitalisation

Predicting the need for hospitalisation of patients with coronavirus disease 2019 (COVID-19) is important for preventing healthcare disruptions. This observational study aimed to use the COVID-19 Registry Japan (COVIREGI-JP) to develop a simple scoring system to predict respiratory failure due to COVID-19 using only underlying diseases and symptoms. A total of 6873 patients with COVID-19 admitted to Japanese medical institutions between 1 June 2020 and 2 December 2020 were included and divided into derivation and validation cohorts according to the date of admission. We used multivariable logistic regression analysis to create a simple risk score model, with respiratory failure as the outcome for young (18-39 years), middle-aged (40-64 years) and older (≥65 years) groups, using sex, age, body mass index, medical history and symptoms. The models selected for each age group were quite different. Areas under the receiver operating characteristic curves for the simple risk score model were 0.87, 0.79 and 0.80 for young, middle-aged and elderly derivation cohorts, and 0.81, 0.80 and 0.67 in the validation cohorts. Calibration of the model was good. The simple scoring system may be useful in the appropriate allocation of medical resources during the COVID-19 pandemic.

Pretransplant Comprehensive Geriatric Assessment in Older Patients with Advanced Chronic Kidney Disease

BACKGROUND

In older patients with advanced chronic kidney disease (CKD), the decision of kidney transplantation (KT) is a challenge for nephrologists. The use of comprehensive geriatric assessment (CGA) is increasingly gaining interest into the process of decision-making about treatment modality choice for CKD. The aim of this study was to assess the prevalence of geriatric impairment and frailty in older dialysis and nondialysis patients with advanced CKD using a pretransplant CGA model and to identify geriatric impairments influencing the geriatricians' recommendations for KT.

METHODS

An observational study was conducted with retrospective data from July 2017 to January 2020. Patients aged ≥65 years with advanced CKD, treated or not with dialysis, and referred by the nephrologist were included in the study. The CGA assessed comorbidity burden, cognition, mood, nutritional status, (instrumental) activities of daily living, physical function, frailty, and polypharmacy. Geriatric impairments influencing the geriatricians' recommendations for KT were identified using univariate and multivariate logistic regressions.

RESULTS

156 patients were included (74.2 ± 3.5 years and 62.2% on dialysis). Geriatric conditions were highly prevalent in both dialysis and nondialysis groups. The rate of geriatric impairments was higher in dialysis patients regarding comorbidity burden, symptoms of depression, physical function, autonomy, and frailty. Geriatrician's recommendations for KT were as follows: favorable (79.5%) versus not favorable or multidisciplinary discussion needed with nephrologists (20.5%). Dependence for Instrumental Activities of Daily Living (IADL) (odds ratio [OR] = 3.01 and 95% confidence interval [CI] = 1.30-7.31), physical functions (OR = 2.91 and 95% CI = 1.08-7.87), and frailty (OR = 2.66 and 95% CI = 1.07-6.65) were found to be independent geriatric impairments influencing geriatrician's recommendations for KT.

CONCLUSIONS

Understanding the burden of geriatric impairment provides an opportunity to direct KT decision-making and to guide interventions to prevent functional decline and preserve quality of life.

Risk Factors and Clinical Outcomes of Cognitive Impairment in Diabetic Patients Undergoing Peritoneal Dialysis

INTRODUCTION

Cognitive impairment (CI) is common in patients with CKD or diabetes mellitus (DM). However, the relevance between DM and CI in diabetic patients undergoing peritoneal dialysis (PD) has not been clearly established. This study aimed to explore the role of DM in CI, the association of glycemic control with CI, and clinical outcomes of CI in diabetic PD patients.

METHODS

Continuous ambulatory PD (CAPD) patients followed up in our PD center between 2014 and 2016 were enrolled. The participants were followed until an endpoint was reached or December 2017. Demographic data and clinical characteristics were collected, and laboratory parameters were measured. The Montreal Cognitive Assessment (MoCA) was used to evaluate global cognitive function, and a score of <26 was considered to indicate CI. A propensity score matching according to age, gender, and mean arterial pressure was conducted between the DM and non-DM groups.

RESULTS

A total of 913 CAPD patients were enrolled, of whom 186 (20.4%) had diabetes. After appropriate matching, 175 patients in the DM group and 270 patients in the non-DM group were included. Patients with diabetes had a higher prevalence of CI and lower scores for visuospatial/executive function, naming, language, delayed recall, and orientation. Higher HbA1c (odds ratio [OR], 1.547; 95% confidence interval [95% CI], 1.013-2.362) and cardiovascular disease (CVD; OR, 2.926; 95% CI, 1.139-7.516) significantly correlated with a risk of CI in diabetic patients. During a median of 26.0 (interquartile range 13.5-35.6) months of follow-up, diabetic patients with CI demonstrated a significantly lower survival rate than those without CI, and CI was an independent risk factor for mortality after adjustment (hazard ratio, 7.224; 95% CI, 1.694-30.806). However, they did not show worse technique survival or higher peritonitis rate than patients without CI.

CONCLUSIONS

HbA1c and CVD are independent risk factors for CI in diabetic patients undergoing CAPD, and CI is independently associated with a higher risk of mortality.

Relationships among the Dosage of Erythropoiesis-Stimulating Agents, Erythropoietin Resistance Index, and Mortality in Maintenance Hemodialysis Patients

BACKGROUND

Erythropoiesis-stimulating agents (ESAs) constitute an important treatment option for anemia in hemodialysis (HD) patients. We investigated the relationships among the dosage of ESA, erythropoietin resistance index (ERI) scores, and mortality in Chinese MHD patients.

METHODS

This multicenter observational retrospective study included MHD patients from 16 blood purification centers (n = 824) who underwent HD in 2011-2015 and were followed up until December 31, 2016. We collected demographic variables, HD parameters, laboratory values, and ESA dosages. Patients were grouped into quartiles according to ESA dosage to study the effect of ESA dosage on all-cause mortality. The ERI was calculated as follows: ESA (IU/week)/weight (kg)/hemoglobin levels (g/dL). We also compared outcomes among the patients stratified into quartiles according to ERI scores. We used the Cox proportional hazards model to measure the relationships between the ESA dosage, ERI scores, and all-cause mortality. Using propensity score matching, we compared mortality between groups according to ERI scores, classified as either > or ≤12.80.

RESULTS

In total, 824 patients were enrolled in the study; 200 (24.3%) all-cause deaths occurred within the observation period. Kaplan-Meier analyses showed that patients administered high dosages of ESAs had significantly worse survival than those administered low dosages of ESAs. A multivariate Cox regression identified that high dosages of ESAs could significantly predict mortality (ESA dosage >10,000.0 IU/week, HR = 1.59, 95% confidence intervals (CIs) (1.04, 2.42), and p = 0.031). Our analysis also indicated a significant increase in the risk of mortality in patients with high ERI scores. Propensity score matching-analyses confirmed that ERI > 12.80 could significantly predict mortality (HR = 1.56, 95% CI [1.11, 2.18], and p = 0.010).

CONCLUSIONS

Our data suggested that ESA dosages >10,000.0 IU/week in the first 3 months constitute an independent predictor of all-cause mortality among Chinese MHD patients. A higher degree of resistance to ESA was related to a higher risk of all-cause mortality.

Application of regularized regression to identify novel predictors of mortality in a cohort of hemodialysis patients

Cohort studies often provide a large array of data on study participants. The techniques of statistical learning can allow an efficient way to analyze large datasets in order to uncover previously unknown, clinically relevant predictors of morbidity or mortality. We applied a combination of elastic net penalized Cox regression and stability selection with the aim of identifying novel predictors of mortality in a cohort of prevalent hemodialysis patients. In our analysis we included 475 patients from the "rISk strAtification in end-stage Renal disease" (ISAR) study, who we split into derivation and confirmation cohorts. A wide array of examinations was available for study participants, resulting in over a hundred potential predictors. In the selection approach many of the well established predictors were retrieved in the derivation cohort. Additionally, the serum levels of IL-12p70 and AST were selected as mortality predictors and confirmed in the withheld subgroup. High IL-12p70 levels were specifically prognostic of infection-related mortality. In summary, we demonstrate an approach how statistical learning can be applied to a cohort study to derive novel hypotheses in a data-driven way. Our results suggest a novel role of IL-12p70 in infection-related mortality, while AST is a promising additional biomarker in patients undergoing hemodialysis.

Effect of comorbidities on survival in patients >80 years of age at onset of renal replacement therapy: data from the ERA-EDTA Registry

BACKGROUND

The number of elderly patients on renal replacement therapy (RRT) is increasing. The survival and quality of life of these patients may be lower if they have multiple comorbidities at the onset of RRT. The aim of this study was to explore whether the effect of comorbidities on survival is similar in elderly RRT patients compared with younger ones.

METHODS

Included were 9333 patients ≥80 years of age and 48 352 patients 20-79 years of age starting RRT between 2010 and 2015 from 15 national or regional registries submitting data to the European Renal Association-European Dialysis and Transplantation Association Registry. Patients were followed until death or the end of 2016. Survival was assessed by Kaplan-Meier curves and the relative risk of death associated with comorbidities was assessed by Cox regression analysis.

RESULTS

Patients ≥80 years of age had a greater comorbidity burden than younger patients. However, relative risks of death associated with all studied comorbidities (diabetes, ischaemic heart disease, chronic heart failure, cerebrovascular disease, peripheral vascular disease and malignancy) were significantly lower in elderly patients compared with younger patients. Also, the increase in absolute mortality rates associated with an increasing number of comorbidities was smaller in elderly patients.

CONCLUSIONS

Comorbidities are common in elderly patients who enter RRT, but the risk of death associated with comorbidities is less than in younger patients. This should be taken into account when assessing the prognosis of elderly RRT patients.

How to relate to dialysis patients' fatigue - perspectives of dialysis nurses and renal health professionals: A qualitative study

BACKGROUND

Although fatigue among the dialysis population is known as a severe and debilitating health problem, this symptom is often under recognized and undertreated.

OBJECTIVE

This qualitative study aimed to gain a better understanding of how dialysis nurses and renal health professionals perceive and address dialysis patient's fatigue in renal care practice.

DESIGN

We conducted a qualitative descriptive study to explore how nurses and renal health professionals perceive and address dialysis patients' fatigue in their daily health care practices. A constructivist grounded theory approach guided analysis and conceptualisation of findings.

SETTING(S)

This study took place at 13 academic and regional settings across the Netherlands. The study was approved by the Medical Ethics Committee (2015.049), on behalf of VU University medical center in Amsterdam.

PARTICIPANTS

Twenty-one renal health professionals of various disciplines took part in interviews: ten dialysis nurses, four nephrologists, two physician assistants, five medical social workers.

METHODS

Semi-structured interviews were conducted in order to gain in-depth insight into the perspectives of dialysis nurses and renal health professionals. An inductive thematic analysis provided insight into health professionals' stances toward dialysis patients' fatigue in light of their daily care context and practices.

RESULTS

Two main themes emerged; 1) 'Fatigue in the background': Shows there is strong focus on medical-technical aspects of the disease. All health professionals perceive fatigue as an intangible symptom that is difficult to address, and falls outside their scope of responsibility and competence. Communication about fatigue among professionals and with patients is limited, pushing fatigue further into the background. 2) 'Vulnerabilities in the background': Especially nurses and social workers signal the accumulating vulnerabilities of dialysis patients and associate these with fatigue (old age, multimorbidities, financial and social problems). Although the need for psychological support is acknowledged, multiple vulnerabilities increase the complexity and intensity of care, and further strengthens the medical-technical focus of care and treatment.

CONCLUSIONS

There is a need to enable renal health professionals to communicate about the complex nature of fatigue in renal patients and stimulate interdisciplinary exchange and shared responsibility. Dialysis nurses have frequent contact with patients during dialysis treatment and are the first to notice when patients' fatigue increases and their overall condition deteriorates. They can play an important role to go beyond the technological imperative of care and understand the lived experiences of patients within their social contexts.

Validation of prognostic indices for short term mortality in an incident dialysis population of older adults >75

Rational and objective

Prognosis provides critical knowledge for shared decision making between patients and clinicians. While several prognostic indices for mortality in dialysis patients have been developed, their performance among elderly patients initiating dialysis is unknown, despite great need for reliable prognostication in that context. To assess the performance of 6 previously validated prognostic indices to predict 3 and/or 6 months mortality in a cohort of elderly incident dialysis patients.

Study design

Validation study of prognostic indices using retrospective cohort data. Indices were compared using the concordance (“c”)-statistic, i.e. area under the receiver operating characteristic curve (ROC). Calibration, sensitivity, specificity, positive and negative predictive values were also calculated.

Setting & participants

Incident elderly (age ≥75 years; n = 349) dialysis patients at a tertiary referral center.

Established predictors

Variables for six validated prognostic indices for short term (3 and 6 month) mortality prediction (Foley, NCI, REIN, updated REIN, Thamer, and Wick) were extracted from the electronic medical record. The indices were individually applied as per each index specifications to predict 3- and/or 6-month mortality.

Results

In our cohort of 349 patients, mean age was 81.5±4.4 years, 66% were male, and median survival was 351 days. The c-statistic for the risk prediction indices ranged from 0.57 to 0.73. Wick ROC 0.73 (0.68, 0.78) and Foley 0.67 (0.61, 0.73) indices performed best. The Foley index was weakly calibrated with poor overall model fit (p <0.01) and overestimated mortality risk, while the Wick index was relatively well-calibrated but underestimated mortality risk.

Limitations

Small sample size, use of secondary data, need for imputation, homogeneous population.

Conclusion

Most predictive indices for mortality performed moderately in our incident dialysis population. The Wick and Foley indices were the best performing, but had issues with under and over calibration. More accurate indices for predicting survival in older patients with kidney failure are needed.

Temporal bone carcinoma: testing the prognostic value of a novel clinical and histological scoring system

PURPOSE

Our group recently proposed the novel Padova prognostic scoring system for temporal bone carcinoma (TBSCC) that considers: the revised Pittsburgh staging system; radiological dura mater involvement; non-anterior spread (medial, inferior or posterior into the temporal bone and skull base) of T4 tumors; and histological grade. The aim of the present study was to validate this prognostic TBSCC scoring system in a case series selected from the literature.

METHODS

A search was run to identify studies on TBSCC reporting the variables included in our score for each patient. Then our system was applied to the data extracted.

RESULTS

Only two published investigations reported all the clinical and pathological data required for our scoring system. In one series from the Gruppo Otologico in Piacenza (Italy), a significantly higher recurrence rate (p = 0.008), shorter disease-free survival (DFS) (p = 0.001), higher disease-specific mortality (DSM) (p = 0.006), and shorter disease-specific survival (DSS) (p = 0.004) were associated with scores ≥ 5. Receiver operating curve (ROC) analysis showed an AUC of 0.804 for TBSCC recurrence, and 0.832 for DSM. In a series from Kyushu University Hospital (Japan), a significantly higher DSM (p = 0.018) and shorter DSS (p = 0.021) were associated with scores ≥ 5. ROC analysis showed an AUC of 0.812 for tumor relapse and 0.790 for DSM.

CONCLUSION

Our TBSCC Padova scoring system confirmed its validity when applied to the only two international TBSCC series providing the required data. These preliminary results need to be confirmed in a multi-center prospective setting.

Elevated Circulating Cell-Free DNA in Hemodialysis-Treated Patients Is Associated with Increased Mortality

BACKGROUND

Predicting the mortality risk of patients un-dergoing hemodialysis (HD) is challenging. Cell-free DNA (cfDNA) is released into circulation from dying cells, and its elevation is predictive of unfavorable outcome. In a pilot study, we found post-HD cfDNA level to be a predictor of all-cause mortality. Thus, the aim of this study was to confirm the prognostic power of cfDNA in a larger prospective cohort study conducted at 2 medical centers.

METHODS

CfDNA levels were measured by a rapid fluorometric assay on sera obtained before and after 1 HD session. One hundred fifty-three patients were followed up to 46 months for mortality during which time 47 patients died. We compared the predictive value of cfDNA to age, comorbidities, and standard blood tests.

RESULTS

Examining standard blood tests, only post-HD cfDNA levels were elevated in the non-survivor group compared to survivors (959 vs. 803 ng/mL, p = 0.04). Pre- and post-HD cfDNA levels correlated with age and diabetes. Patients with elevated cfDNA (>850 ng/mL) showed lower survival than those with normal levels. A Cox proportional hazard regression model demonstrated a significant hazard ratio of 1.92 for post-HD cfDNA levels. Logistic regression models showed that post-HD cfDNA was a significant predictor of mortality at 1-3 years with odd ratios of 4.61, 4.36, and 6.22, respectively.

CONCLUSIONS

Post-HD cfDNA level was superior to standard blood tests and could serve as a biomarker to assist in decision-making for HD-treated patients.

Integrating the Surprise Question, Palliative Care Screening Tool, and Clinical Risk Models to Identify Peritoneal Dialysis Patients With High One-Year Mortality

CONTEXT

Universal screening to identify vulnerable patients who may receive limited benefits from life-sustaining treatments can facilitate palliative care in dialysis populations.

OBJECTIVES

We aimed to develop prediction models for one-year mortality in peritoneal dialysis (PD) patients.

METHODS

This prospective cohort study included 401 adult Taiwanese prevalent PD patients (average age 56.2 ± 14 years). In addition to obtaining clinical characteristics and laboratory data, the primary care nurses evaluated the surprise question (SQ) and palliative care screening tool (PCST) for each patient in March 2015. Multivariate logistic regression models were conducted to predict the primary outcome of one-year all-cause mortality.

RESULTS

There were 34 (8.5%) patients who died during the first year of follow-up. Patients allocated to the not surprised group according to the SQ and those who received a score of ≥4 on the PCST had increased odds of death (odds ratio 24.68 [95% CI 10.66-57.13] and 12.18 [95% CI 5.66-26.21], respectively). We also developed a clinical risk model for one-year mortality that included sex, dialysis vintage, coronary artery disease, malignancy, normalized protein nitrogen appearance, white blood cell count, and serum albumin and sodium levels. Integrating the SQ, PCST, and clinical risk model exhibited good discrimination with an area under the receiver operating characteristic curve of 0.95. Kaplan-Meier analysis showed worse survival in high-risk patients predicted by the integrated model (log-rank P < 0.001).

CONCLUSION

Screening with the use of the integrated measurement can identify high-risk PD patients. This approach may facilitate palliative care interventions for at-risk subpopulations.

The current and future landscape of dialysis

The development of dialysis by early pioneers such as Willem Kolff and Belding Scribner set in motion several dramatic changes in the epidemiology, economics and ethical frameworks for the treatment of kidney failure. However, despite a rapid expansion in the provision of dialysis — particularly haemodialysis and most notably in high-income countries (HICs) — the rate of true patient-centred innovation has slowed. Current trends are particularly concerning from a global perspective: current costs are not sustainable, even for HICs, and globally, most people who develop kidney failure forego treatment, resulting in millions of deaths every year. Thus, there is an urgent need to develop new approaches and dialysis modalities that are cost-effective, accessible and offer improved patient outcomes. Nephrology researchers are increasingly engaging with patients to determine their priorities for meaningful outcomes that should be used to measure progress. The overarching message from this engagement is that while patients value longevity, reducing symptom burden and achieving maximal functional and social rehabilitation are prioritized more highly. In response, patients, payors, regulators and health-care systems are increasingly demanding improved value, which can only come about through true patient-centred innovation that supports high-quality, high-value care. Substantial efforts are now underway to support requisite transformative changes. These efforts need to be catalysed, promoted and fostered through international collaboration and harmonization.

Dialysis is a life-saving therapy; however, costs of dialysis are high, access is inequitable and outcomes are inadequate. This Review describes the current landscape of dialysis therapy from an epidemiological, economic, ethical and patient-centred framework, and describes initiatives that are aimed at stimulating innovations in the field to one that supports high-quality, high-value care.

The global dialysis population is growing rapidly, especially in low-income and middle-income countries; however, worldwide, a substantial number of people lack access to kidney replacement therapy, and millions of people die of kidney failure each year, often without supportive care.

The costs of dialysis care are high and will likely continue to rise as a result of increased life expectancy and improved therapies for causes of kidney failure such as diabetes mellitus and cardiovascular disease.

Patients on dialysis continue to bear a high burden of disease, shortened life expectancy and report a high symptom burden and a low health-related quality of life.

Patient-focused research has identified fatigue, insomnia, cramps, depression, anxiety and frustration as key symptoms contributing to unsatisfactory outcomes for patients on dialysis.

Initiatives to transform dialysis outcomes for patients require both top-down efforts (that is, efforts that promote incentives based on systems level policy, regulations, macroeconomic and organizational changes) and bottom-up efforts (that is, patient-led and patient-centred advocacy efforts as well as efforts led by individual teams of innovators).

Patients, payors, regulators and health-care systems increasingly demand improved value in dialysis care, which can only come about through true patient-centred innovation that supports high-quality, high-value care.

Physical Function Assessment in Older Hemodialysis Patients

RATIONALE & OBJECTIVE

Physical function is not routinely measured in older adults receiving dialysis. We evaluated the appropriateness of repeated measurements of physical function, including Short Physical Performance Battery (SPPB), handgrip strength, and activities of daily living (ADLs), in older adults receiving dialysis.

STUDY DESIGN

Prospective study.

SETTING & PARTICIPANTS

37 community-dwelling adults 65 years and older receiving in-center hemodialysis at 5 dialysis units located in North Carolina.

EXPOSURES

SPPB (an assessment of standing balance, chair stands, and gait speed), handgrip strength, and Katz and Lawton ADLs at baseline and subsequent 3-month intervals up to 6 months.

OUTCOMES

Completion rate, presence of floor or ceiling effects, and presence of clinically meaningful change in physical function measurements.

RESULTS

Of 55 potential participants, we enrolled 37 (67%) older adults receiving hemodialysis. Among 35 enrolled participants who completed baseline assessment in a dialysis unit, mean age was 70.1 (SD, 5) years, 46% (n = 16) were women, 77% (n = 27) were African American, and median time receiving dialysis was 2.7 (IQR, 0.6-5.0) years. There were 3 deaths within the observation period, and study retention at 3 and 6 months was 83% (n = 29) and 74% (n = 26), respectively. Participants tolerated measurements; only 2 participants did not attempt 1 of the performance-based tests at a study visit. Baseline median SPPB score, grip strength, and gait speed were 6 (IQR, 4-9), 55 (IQR, 42-70) kg, and 0.76 (IQR, 0.46-0.86) m/s, respectively. Baseline median for Katz and Lawton ADLs were 6 (IQR, 6-6) and 7 (IQR, 4-8), respectively; ceiling effects were observed for both measures. For some participants, clinically meaningful changes (improvement or decline) in SPPB score, grip strength, and gait speed occurred at each 3-month interval.

LIMITATIONS

Limited geographic and ethnic variation.

CONCLUSIONS

SPPB, handgrip strength, and gait speed alone are appropriate measures for interval physical function assessment in community-dwelling older adults receiving in-center hemodialysis.

Exploring the choices and outcomes of older patients with advanced kidney disease

A lack of data on patient choices and outcomes at the time of pre-dialysis planning limits meaningful shared decision making, particularly in older frailer patients. In this large retrospective cohort study of patients aged over 70 seen by the pre-dialysis clinic (2004–2016) of a large single centre in the United Kingdom (1,216 patients), age, sex, comorbidity, poverty and frailty were used to predict choice of renal replacement therapy (RRT) over maximum conservative management (MCM). The impact of patient choice of RRT versus MCM was used to predict survival from the time of choice using multivariable Cox proportional hazards regression. Older age, female sex, greater poverty and greater frailty were associated with choosing MCM, whilst comorbidity had no significant impact on choice. At 5 years of follow up, 49% of all patients had died without receiving RRT. Over 70% of the patients choosing MCM died with better kidney function than the median level at which those starting RRT initiated treatment. Frailty and age were better predictors of survival than comorbidity and in patients with at least moderate frailty, RRT offered no survival benefit over MCM. In conclusion, analysing outcomes from the time of choice may improve shared decision making. Frailty should be routinely assessed and collected and further work may help predict which patients are unlikely to survive or progress to end stage renal disease and may not need to be burdened with making a pre-dialysis choice.

Contribution of medico-administrative data to the development of a comorbidity score to predict mortality in End-Stage Renal Disease patients

Comorbidity scores to predict mortality are very useful to facilitate decision-making for personalized patient management. This study aim was to assess the contribution of medico-administrative data in addition to French Renal Epidemiology and Information Network (REIN) data to the development of a risk score to predict the 1-year all-cause mortality in patients with End Stage Renal Disease (ESRD), and to compare it with previous scores. Data from a derivation sample (n = 6336 patients who started dialysis in 2015 in France) obtained by linking the REIN and the French National Health Insurance Information System databases were analyzed with multivariate Cox models to select risk factors to establish the score. A randomly chosen validation sample (n = 2716 patients who started dialysis in 2015) was used to validate the score and to compare it with the comorbidity indexes developed by Wright and Charlson. The ability to predict one-year mortality of the score constructed using REIN data linked to the medico-administrative database was not higher than that of the score constructed using only REIN data (i.e., Rennes score). The Rennes score included five comorbidities, albumin, and age. This score (AUC = 0.794, 95%CI: 0.768–0.821) outperformed both the Wright (AUC = 0.631, 95%CI: 0.621–0.639; p < 0.001) and Charlson (AUC = 0.703, 95%CI: 0.689–0.716; p < 0.001) indexes. Data from the REIN registry alone, collected at dialysis start, are sufficient to develop a risk score that can predict the one-year mortality in patients with ESRD. This simple score might help identifying high risk patients and proposing the most adapted care.

In-hospital mortality in elderly patients with acute kidney injury requiring dialysis: a cohort analysis

OBJECTIVE

To determine risk factors for in-hospital mortality in elderly patients with acute kidney injury (AKI) requiring dialysis.

INTRODUCTION

AKI requiring dialysis is frequent in elderly and is associated with an increased intra-hospital mortality. With the growing number of older individuals among hospitalized patients with AKI demands a thorough investigation of the factors that contribute to their mortality to improve outcomes.

METHODS

We performed a retrospective analysis of patients older than 80 years, admitted due to AKI requiring dialysis between January 2016 and December 2017. Patients who need intensive-care units (ICU) admission were excluded. The primary outcome was all-cause in-hospital mortality.

RESULTS

A total of 154 patients were evaluated. The mean age was 85.3 ± 4.0 years and 76 patients (49.4%) were male. The overall mortality rate was 26.6%. On the multivariate analysis, serum albumin (OR 0.42 [95% CI 0.21-0.85], p 0.016), C reactive protein/albumin ratio (OR 1.04 [95% CI 0.99-1.09], and renal function recovery (OR 018 [95% CI 0.49-0.65], p 0.009) were the factors associated with higher in-hospital mortality.

CONCLUSIONS

Lower albumin level, higher C reactive protein/albumin ratio at admission, and absence of renal function recovery are associated with increased in-hospital mortality's risk in elderly with acute kidney injury requiring dialysis.

Patient-nephrologist prognostic awareness and discordance in end stage renal disease on renal replacement therapy

PURPOSE

The 1-year mortality rate of patients with end-stage renal disease (ESRD) on renal replacement therapy (RRT) is 20-25% comparable to many cancers. Studies have shown that cancer patients commonly overestimate their likelihood of survival relative to their physicians. It is unclear if this translates into other terminal illnesses.

METHODS

Adult and elderly patients with ESRD on RRT without cognitive defect were interviewed to evaluate their prognostic estimates at 1 and 5 years. Past medical history and demographic data was abstracted from their medical charts. Each patient's proper nephrologist was then interviewed regarding his proper prognostic estimate for this patient. Both the patient and the nephrologist's estimates were compared and a difference of greater than 20% was defined as the threshold for prognostic concordance.

RESULTS

77% of patients were found to be in prognostic discordance with their nephrologists. This group was older, had more comorbidities, a lower albumin level and a poorer prognosis. The majority of patients were in disagreement with their nephrologists regarding whether a discussion about prognosis had taken place. The choice of end of life care for 55% of patients was focused on relieving pain and discomfort.

CONCLUSION

Communication of prognosis and discussions related to life expectancy and end of life care are lacking in the routine care of ESRD patients. ESRD patients therefore tend to overestimate their prognosis which might lead to overutilization of invasive procedures with increased acute healthcare costs as well as a delay in instigation of palliative or hospice measures.

Enabling good outcomes in older adults on dialysis: a qualitative study

Background

Older patients on dialysis may not have optimal outcomes, particularly with regards to quality of life. Existing research is focused mainly on survival, with limited information about other outcomes. Such information can help in shared decision-making around dialysis initiation; it can also be used to improve outcomes in patients established on dialysis. We used qualitative research methods to explore patient perspectives regarding their experience and outcomes with dialysis.

Methods

Semi-structured interviews with participants aged ≥70, receiving dialysis at a regional Australian hospital, were recorded and transcribed. From participants’ responses, we identified descriptive themes using a phenomenological approach, with verification by two researchers. Factors affecting outcomes were derived reflexively from these themes.

Results

Seventeen interviews were analysed prior to saturation of themes. Participants (12 on haemodialysis, 5 on peritoneal dialysis) had spent an average of 4.3 years on dialysis. There were 11 males and 6 females, with mean age 76.2 years (range 70 to 83). Experiences of dialysis were described across four domains - the self, the body, effects on daily life and the influences of others; yielding themes of (i) responses to loss (of time, autonomy, previous life), (ii) responses to uncertainty (variable symptoms; unpredictable future; dependence on others), (iii) acceptance / adaptation (to life on dialysis; to ageing) and (iv) the role of relationships / support (family, friends and clinicians).

Conclusions

Older patients experience the effects of dialysis across multiple domains in their lives. They endure feelings of loss and persistent uncertainty, but may also adapt successfully to their new circumstances, aided by the support they receive from family, health professionals and institutions. From these insights, we have suggested practical measures to improve outcomes in older patients.

Predicting 6-Month Mortality in Incident Elderly Dialysis Patients: A Simple Prognostic Score

AIM

Mortality in end-stage renal disease (ESRD) remains high, particularly among elderly, who represents the most rapidly growing segment of the ESRD population in wealthier countries. We developed and validated a risk score in elderly patients to predict 6-month mortality after dialysis initiation.

METHODS

We used data from a cohort of 421 patients, aged 65 years and over who started dialysis between 2009 and 2016, in our Nephrology department. The predictive score was developed using a multivariable logistic regression analysis. A bootstrapping technique was used for internal validation.

RESULTS

The overall mortality within 6 months was 14.0%. Five independent predictors were identified, and a points system was constructed: age 75 years or older (2 points), coronary artery disease (2), cerebrovascular disease with hemiplegia (2), time of nephrology care before dialysis (<3.0 months [2]; ≥3 to <12 months [1]), and serum albumin levels (3.0-3.49 g/dL [1]; <3.0 g/dL [2]). A score of 6 identified patients with a 70% risk of 6-month mortality. Model performance was good in both discrimination (area under the curve of 0.793; [95% CI 0.73-0.86]) and validation (concordance statistics of 0.791 [95% CI 0.73-0.85]).

CONCLUSIONS

We developed a simple prediction score based on readily available clinical and laboratory data that can be a practical and useful tool to assess short-term prognosis in elderly patients starting dialysis. It may help to inform patients and their families about ESRD treatment options and provide a more patient-centered overall approach to care.

Discussions during shared decision-making in older adults with advanced renal disease: a scoping review

OBJECTIVES

This review summarises the information available for clinicians counselling older patients with kidney failure about treatment options, focusing on prognosis, quality of life, the lived experiences of treatment and the information needs of older adults.

DESIGN

We followed the Joanna Briggs Institute Methodology for Scoping Reviews. The final report conforms to the PRISMA-ScR guidelines.

DATA SOURCES

PubMed, PsycINFO, CINAHL, Embase, Scopus, Web of Science, TRIP and online repositories (for dissertations, guidelines and recommendations from national renal associations).

ELIGIBILITY CRITERIA FOR INCLUSION

Articles in English studying older adults with advanced kidney disease (estimated glomerular filtration rate <30 mL/min/1.73 m²); published between January 2000 and August 2018. Articles not addressing older patients separately or those comparing between dialysis modalities were excluded.

DATA EXTRACTION AND SYNTHESIS

Two independent reviewers screened articles for inclusion and grouped them by topic as per the objectives above. Quantitative data were presented as tables and charts; qualitative themes were identified and described.

RESULTS

248 articles were included after screening 15 445 initial results. We summarised prognostic scores and compared dialysis and non-dialytic care. We highlighted potentially modifiable factors affecting quality of life. From reports of the lived experiences, we documented the effects of symptoms, of ageing, the feelings of disempowerment and the need for adaptation. Exploration of information needs suggested that patients want to participate in decision-making and need information, in simple terms, about survival and non-survival outcomes.

CONCLUSION

When discussing treatment options, validated prognostic scores are useful. Older patients with multiple comorbidities do not do well with dialysis. The modifiable factors contributing to the low quality of life in this cohort deserve attention. Older patients suffer a high symptom burden and functional deterioration; they have to cope with significant life changes and feelings of disempowerment. They desire greater involvement and more information about illness, symptoms and what to expect with treatment.

Medical Record Documentation of Goals-of-Care Discussions Among Older Veterans With Incident Kidney Failure

RATIONALE & OBJECTIVE

Elicitation and documentation of patient preferences is at the core of shared decision making and is particularly important among patients with high anticipated mortality. The extent to which older patients with incident kidney failure undertake such discussions with their providers is unknown and its characterization was the focus of this study.

STUDY DESIGN

Retrospective cohort study.

SETTING & PARTICIPANTS

A random sample of veterans 67 years and older with incident kidney failure receiving care from the US Veterans Health Administration between 2005 and 2010.

EXPOSURES

Demographic and facility characteristics, as well as predicted 6-month mortality risk after dialysis initiation and documentation of resuscitation preferences.

OUTCOMES

Documented discussions of dialysis treatment and supportive care.

ANALYTICAL APPROACH

We reviewed medical records over the 2 years before incident kidney failure and up to 1 year afterward to ascertain the frequency and timing of documented discussions about dialysis treatment, supportive care, and resuscitation. Logistic regression was used to identify factors associated with these documented discussions.

RESULTS

The cohort of 821 veterans had a mean age of 80.9±7.2 years, and 37.2% had a predicted 6-month mortality risk>20% with dialysis. Documented discussions addressing dialysis treatment and resuscitation were present in 55.6% and 77.1% of patients, respectively. Those addressing supportive care were present in 32.4%. The frequency of documentation varied by mortality risk and whether the patient ultimately started dialysis. In adjusted analyses, the frequency and pattern of documentation were more strongly associated with geographic location and receipt of outpatient nephrology care than with patient demographic or clinical characteristics.

LIMITATIONS

Documentation may not fully reflect the quality and content of discussions, and generalizability to nonveteran patients is limited.

CONCLUSIONS

Among older veterans with incident kidney failure, discussions of dialysis treatment are decoupled from other aspects of advance care planning and are suboptimally documented, even among patients at high risk for mortality.

Prediction model for cardiovascular events or all-cause mortality in incident dialysis patients

Some variables including age, comorbidity of diabetes, and so on at dialysis initiation are associated with patient prognosis. Cardiovascular (CV) events are a major cause of death, and adequate models that predict prognosis in dialysis patients are warranted. Therefore, we created models using some variables at dialysis initiation. We used a database of 1,520 consecutive dialysis patients (median age, 70 years; 492 women [32.4%]) from a multicenter prospective cohort study. We established the primary endpoint as a composite of the incidence of first CV events or all-cause death. A multivariable Cox proportional hazard regression model was used to construct a model. We considered a complex and a simple model. We used area under the receiver operating characteristic curve (AUROC) to assess and compare the predictive performances of the prediction models and evaluated the improvement in discrimination using the complex model versus the simple model using net reclassification improvement (NRI). We then assessed integrated discrimination improvement (IDI) to evaluate improvements in average sensitivity and specificity. Of 392 deaths, 152 were CV-related. Totally, 506 CV events occurred during the follow-up period (median 1,285 days). Finally, 692 patients reached the primary endpoint. Baseline data were set at dialysis initiation. AUROC for the primary endpoint was 0.737 (95% confidence interval [CI], 0.712–0.761) in the simple model and 0.765 (95% CI, 0.741–0.788) in the complex model. There were significant intergroup differences in NRI (0.44; 95% CI, 0.34–0.53; p < 0.001) and IDI (0.02; 95% CI, 0.02–0.03; p < 0.001). We prepared a Shiny R application for each model to automatically calculate the predicted occurrence probability (https://statacademy.shinyapps.io/App_inaguma_20190717/). The complex model made more accurate predictions than the simple model. However, the intergroup difference was not significant. Hence, the simple model was more useful than the complex model. The tool was useful in a real-world clinical setting because it required only routinely available variables. Moreover, we emphasized that the tool could predict the incidence of CV events or all-cause mortality for individual patients. In the future, we must confirm its external validity in other prospective cohorts.

Hemodialysis vascular access in the elderly-getting it right

Choosing the optimal hemodialysis vascular access for the elderly patient is best achieved by a patient-centered coordinated multidisciplinary team approach that aligns the patient's end-stage kidney disease Life-Plan, i.e., the individual treatment approach (supportive care, time-limited or long-term kidney replacement therapy, or combination thereof) and selection of dialysis modality (peritoneal dialysis versus hemodialysis) with the most suitable dialysis access. Finding the right balance between the patient's preferences, the likelihood of access function and survival, and potential complications in the context of available resources and limited patient survival can be extremely challenging. The framework for choosing the most appropriate vascular access for the elderly presented in this review considers the individual end-stage kidney disease Life-Plan, the patient life expectancy, the likelihood of access function and survival, the timing of dialysis relative to access placement, prior access history, and patient preference. This complex decision-making process should be dynamic in order to accommodate patients' changing needs and life and health circumstances. Effective and timely communication between the patient, their caregivers, and treating team is key to delivering truly patient-centered care. Delivering this care also requires overcoming the limitations of the currently available evidence that is predominantly based on observational data with its inherent risks of bias. While challenging, future randomized controlled studies exploring the risks, benefits, costs, and timing of placement of available access types in the elderly are required to help us "get it right" for our patients.

Geriatric Assessment and the Relation with Mortality and Hospitalizations in Older Patients Starting Dialysis

BACKGROUND AND OBJECTIVES

A geriatric assessment (GA) is a structural method for identifying frail patients. The relation of GA findings and risk of death in end-stage kidney disease (ESKD) is not known. The objective of the GA in OLder patients starting Dialysis Study was to assess the association of GA at dialysis initiation with early mortality and hospitalization.

DESIGN, SETTING, PARTICIPANTS, AND MEASUREMENTS

Patients ≥65 years old were included just prior to dialysis initiation. All participants underwent a GA, including assessment of (instrumental) activities of daily living (ADL), mobility, cognition, mood, nutrition, and comorbidity. In addition, a frailty screening (Fried Frailty Index, [FFI]) was applied. Outcome measures were 6- and 12-month mortality, and 6-month hospitalization. Associations with mortality were assessed with cox-regression adjusting for age, sex, comorbidity burden, smoking, residual kidney function and dialysis modality. Associations with hospitalization were assessed with logistic regression, adjusting for relevant confounders.

RESULTS

In all, 192 patients were included, mean age 75 ± 7 years, of whom 48% had ≥3 geriatric impairments and were considered frail. The FFI screening resulted in 46% frail patients. Mortality rate was 8 and 15% at 6- and 12-months after enrolment, and transplantation rate was 2 and 4% respectively. Twelve-month mortality risk was higher in patients with ≥3 impairments (hazard ratio [HR] 2.97 [95% CI 1.19-7.45]) compared to less impaired patients. FFI frail patients had a higher risk of 12-month mortality (HR 7.22 [95% CI 2.47-21.13]) and hospitalization (OR 1.93 [95% CI 1.00-3.72]) compared to fit patients. Malnutrition was associated with 12-month mortality, while impaired ADL and depressive symptoms were associated with 12-month mortality and hospitalization.

CONCLUSIONS

Frailty as assessed by a GA is related to mortality in elderly patients with ESKD. Individual components of the GA are related to both mortality and hospitalization. As the GA allows for distinguishing between frail and fit patients initiating dialysis, it is potentially of added value in the decision-making process concerning dialysis initiation.

Prediction of Risk of Death for Patients Starting Dialysis: A Systematic Review and Meta-Analysis

BACKGROUND AND OBJECTIVES

Dialysis is a preference-sensitive decision where prognosis may play an important role. Although patients desire risk prediction, nephrologists are wary of sharing this information. We reviewed the performance of prognostic indices for patients starting dialysis to facilitate bedside translation.

DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS

Systematic review and meta-analysis following the PRISMA guidelines. We searched Ovid MEDLINE, Ovid Embase, Ovid Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus for eligible studies of patients starting dialysis published from inception to December 31, 2018.

SELECTION CRITERIA

Articles describing validated prognostic indices predicting mortality at the start of dialysis. We excluded studies limited to prevalent dialysis patients, AKI and studies excluding mortality in the first 1-3 months. Two reviewers independently screened abstracts, performed full text assessment of inclusion criteria and extracted: study design, setting, population demographics, index performance and risk of bias. Pre-planned random effects meta-analysis was performed stratified by index and predictive window to reduce heterogeneity.

RESULTS

Of 12,132 articles screened and 214 reviewed in full text, 36 studies were included describing 32 prognostic indices. Predictive windows ranged from 3 months to 10 years, cohort sizes from 46 to 52,796. Meta-analysis showed discrimination area under the curve (AUC) of 0.71 (95% confidence interval, 0.69 to 073) with high heterogeneity (I2=99.12). Meta-analysis by index showed highest AUC for The Obi, Ivory, and Charlson comorbidity index (CCI)=0.74, also CCI was the most commonly used (ten studies). Other commonly used indices were Kahn-Wright index (eight studies, AUC 0.68), Hemmelgarn modification of the CCI (six studies, AUC 0.66) and REIN index (five studies, AUC 0.69). Of the indices, ten have been validated externally, 16 internally and nine were pre-existing validated indices. Limitations include heterogeneity and exclusion of large cohort studies in prevalent patients.

CONCLUSIONS

Several well validated indices with good discrimination are available for predicting survival at dialysis start.

Artificial Intelligence Prediction Model for the Cost and Mortality of Renal Replacement Therapy in Aged and Super-Aged Populations in Taiwan

BACKGROUND

Prognosis of the aged population requiring maintenance dialysis has been reportedly poor. We aimed to develop prediction models for one-year cost and one-year mortality in aged individuals requiring dialysis to assist decision-making for deciding whether aged people should receive dialysis or not.

METHODS

We used data from the National Health Insurance Research Database (NHIRD). We identified patients first enrolled in the NHIRD from 2000-2011 for end-stage renal disease (ESRD) who underwent regular dialysis. A total of 48,153 Patients with ESRD aged ≥65 years with complete age and sex information were included in the ESRD cohort. The total medical cost per patient (measured in US dollars) within one year after ESRD diagnosis was our study's main outcome variable. We were also concerned with mortality as another outcome. In this study, we compared the performance of the random forest prediction model and of the artificial neural network prediction model for predicting patient cost and mortality.

RESULTS

In the cost regression model, the random forest model outperforms the artificial neural network according to the mean squared error and mean absolute error. In the mortality classification model, the receiver operating characteristic (ROC) curves of both models were significantly better than the null hypothesis area of 0.5, and random forest model outperformed the artificial neural network. Random forest model outperforms the artificial neural network models achieved similar performance in the test set across all data.

CONCLUSIONS

Applying artificial intelligence modeling could help to provide reliable information about one-year outcomes following dialysis in the aged and super-aged populations; those with cancer, alcohol-related disease, stroke, chronic obstructive pulmonary disease (COPD), previous hip fracture, osteoporosis, dementia, and previous respiratory failure had higher medical costs and a high mortality rate.

Association of Initiation of Maintenance Dialysis with Functional Status and Caregiver Burden

BACKGROUND AND OBJECTIVES

Little is known about the functional course after initiating dialysis in elderly patients with ESKD. The aim of this study was to assess the association of the initiation of dialysis in an elderly population with functional status and caregiver burden.

DESIGN, SETTING, PARTICIPANTS & MEASUREMENTS

This study included participants aged ≥65 years with ESKD who were enrolled in the Geriatric Assessment in Older Patients Starting Dialysis study. All underwent a geriatric assessment and a frailty screening (Fried Frailty Index and Groningen Frailty Indicator) at dialysis initiation. Functional status (activities of daily life and instrumental activities of daily life) and caregiver burden were assessed at baseline and after 6 months. Decline was defined as loss of one or more domains in functional status, stable as no difference between baseline and follow-up, and improvement as gain of one or more domains in functional status. Logistic regression was performed to assess the association between the combined outcome functional decline/death and potential risk factors.

RESULTS

Of the 196 included participants functional data were available for 187 participants. Mean age was 75±7 years and 33% were women. At the start of dialysis, 79% were care dependent in functional status. After 6 months, 40% experienced a decline in functional status, 34% remained stable, 18% improved, and 8% died. The prevalence of high caregiver burden increased from 23%-38% (P=0.004). In the multivariable analysis age (odds ratio, 1.05; 95% confidence interval, 1.00 to 1.10 per year older at baseline) and a high Groningen Frailty Indicator compared with low score (odds ratio, 1.97; 95% confidence interval, 1.05 to 3.68) were associated with functional decline/death.

CONCLUSIONS

In patients aged ≥65 years, functional decline within the first 6 months after initiating dialysis is highly prevalent. The risk is higher in older and frail patients. Loss in functional status was mainly driven by decline in instrumental activities of daily life. Moreover, initiation of dialysis is accompanied by an increase in caregiver burden.

Machine Learning to Identify Dialysis Patients at High Death Risk

Introduction

Given the high mortality rate within the first year of dialysis initiation, an accurate estimation of postdialysis mortality could help patients and clinicians in decision making about initiation of dialysis. We aimed to use machine learning (ML) by incorporating complex information from electronic health records to predict patients at risk for postdialysis short-term mortality.

Methods

This study was carried out on a contemporary cohort of 27,615 US veterans with incident end-stage renal disease (ESRD). We implemented a random forest method on 49 variables obtained before dialysis transition to predict outcomes of 30-, 90-, 180-, and 365-day all-cause mortality after dialysis initiation.

Results

The mean (±SD) age of our cohort was 68.7 ± 11.2 years, 98.1% of patients were men, 29.4% were African American, and 71.4% were diabetic. The final random forest model provided C-statistics (95% confidence intervals) of 0.7185 (0.6994–0.7377), 0.7446 (0.7346–0.7546), 0.7504 (0.7425–0.7583), and 0.7488 (0.7421–0.7554) for predicting risk of death within the 4 different time windows. The models showed good internal validity and replicated well in patients with various demographic and clinical characteristics and provided similar or better performance compared with other ML algorithms. Results may not be generalizable to non-veterans. Use of predictors available in electronic medical records has limited the assessment of number of predictors.

Conclusion

We implemented and ML-based method to accurately predict short-term postdialysis mortality in patients with incident ESRD. Our models could aid patients and clinicians in better decision making about the best course of action in patients approaching ESRD.